Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2025-11-14
19:53:40 melwitt dansmith: sorry if you already saw but just in case you didn't, I replied to your latest comments on the vtpm stack
22:20:57 opendevreview Merged openstack/nova master: add functional repoducer for bug 2048837 https://review.opendev.org/c/openstack/nova/+/919979
22:26:17 dansmith melwitt: ack, sorry I haven't even looked to see if you had
22:27:18 melwitt oh ok
#openstack-nova - 2025-11-15
10:54:42 opendevreview Balazs Gibizer proposed openstack/nova master: WIP:Move eventlet specific libvirt event handling https://review.opendev.org/c/openstack/nova/+/967178
#openstack-nova - 2025-11-16
05:13:19 opendevreview Ghanshyam proposed openstack/nova master: WIP: Graceful shutodwn of nova services https://review.opendev.org/c/openstack/nova/+/967261
#openstack-nova - 2025-11-17
08:16:44 opendevreview Konrad Gube proposed openstack/nova-specs master: Propose assisted-volume-extend API https://review.opendev.org/c/openstack/nova-specs/+/949504
08:45:13 gibi dansmith: thanks for the review on https://review.opendev.org/c/openstack/nova/+/966446/1#message-d56cbd036e036b8a6962e6023c1c4213fbadf7ab I replied inline
10:33:05 opendevreview Henry Richter proposed openstack/nova master: Preserve vTPM state between power off and power on https://review.opendev.org/c/openstack/nova/+/955657
10:34:51 opendevreview Henry Richter proposed openstack/nova master: Preserve vTPM state between power off and power on https://review.opendev.org/c/openstack/nova/+/955657
10:54:47 opendevreview sean mooney proposed openstack/nova master: ensure correct cleanup of multi-attach volumes https://review.opendev.org/c/openstack/nova/+/916322
11:29:54 LarsErikP tobias-urdin: do you know anything about the status of merging this? https://review.opendev.org/c/openstack/nova/+/916409
13:27:41 opendevreview sean mooney proposed openstack/nova master: ensure correct cleanup of multi-attach volumes https://review.opendev.org/c/openstack/nova/+/916322
13:28:38 sean-k-mooney gibi: good catch i missed that i reordered it ^ shoudl adress your comments
13:35:32 opendevreview Henry Richter proposed openstack/nova master: Preserve vTPM state between power off and power on https://review.opendev.org/c/openstack/nova/+/955657
13:37:23 opendevreview Henry Richter proposed openstack/nova master: Preserve vTPM state between power off and power on https://review.opendev.org/c/openstack/nova/+/955657
13:38:59 gibi sean-k-mooney: thanks. +2
13:47:20 sean-k-mooney long term ther is a more generic version fo the FairLockGuard concept that i may eventually upstream to oslo. if i do write that ill eventurally refactort nova to use it but i want thsi simpler version in nova so we can backport the fix. gibi the more generic version would take to sets of lock names to aquire as readers and writers, this would allow me to create a more
13:47:22 sean-k-mooney generic form of the @serial decorator that we use for tempest so that we can declare annotate test that need exclusive access to a subset of resouces with a request for the wright lock for that resouce, the use case is watcher senario tests have some tests that exclsilve operatate on voluem and ohter that exlucsitlvy operato on instnaces and some that work on both
13:47:42 opendevreview Rajesh Tailor proposed openstack/nova master: Clear image_ref for boot-from-volume instances https://review.opendev.org/c/openstack/nova/+/967369
13:47:56 sean-k-mooney today we run the warcher tempest test with concurrancy 1 so that they dont interfer
13:48:20 sean-k-mooney evecutnally i think we can do better with an enhanced version fo this but that is over kill for what nova needs today
14:27:51 gibi sean-k-mooney: could you check this small hacking rule https://review.opendev.org/c/openstack/nova/+/958408 ?
14:33:56 sean-k-mooney sure
14:36:07 sean-k-mooney ok so we are trying to block import or direct use via fully qulaifed name of eventlets semaphore timeout or event
14:37:01 sean-k-mooney :) and it found one incorrect usage in a test
14:37:55 sean-k-mooney i need to check but we have already converted LibvirtDriver._handle_conn_event to actully use ThreadingEventWithResult() right?
14:39:30 gibi https://github.com/openstack/nova/blob/8a993d583ffedb608a67acac9fce0c3d065e2dd2/nova/compute/manager.py#L268
14:39:33 gibi yepp
14:41:45 sean-k-mooney _InstanceEvents = ty.Dict[ty.Tuple[str, str], ThreadingEventWithResult] ah ok then yes i think this all looks fine to me
14:42:22 sean-k-mooney gibi: i didnt have a chance to look last week but you mentioend the first copule of patches in the eventlet serise are ready to review
14:43:27 sean-k-mooney ah found the base https://review.opendev.org/c/openstack/nova/+/966443/1 looks like stephen has alrady taken a look at that one at least
14:44:11 sean-k-mooney i have maybe 30-40 mins before my next meeting ill see if i can take a look at one or two of them
15:08:46 gibi thanks
15:27:41 dansmith gibi: I replied... I'm concerned we have such differing views of what init_host() is for
15:28:04 dansmith but, I +W'd as I don't think I'm going to convince you and a fixed bug is better of course
15:28:53 dansmith I would offer to tighten up the definition of init_host() but I'd do so in a way incompatible with your definition, so probably better for me to not
15:34:27 gibi dansmith: thanks. "Initialize anything that is necessary for the driver to function" for me means do not use the driver until this is done. I understand that technically we could use parts of the driver already after host.initialize() like querying the uuids. But what you suggest would actually require the better part of the current libvirt/driver.init_host() to be move to __init__ not just
15:34:33 gibi host.initialize() like the libvirt version check. I really don't want to query anything about libvirt before we are sure we are talking about a libvirt we support
15:36:32 gibi and really changing the meaning of init_host during a bugfix seems dangerous
15:36:36 dansmith except we call host.initialize() before we check the libvirt version :)
15:36:59 gibi we do but we don't return from init_host until we check the libvirt version
15:37:00 dansmith well, as I noted, I don't think the current meaning agrees with what you think it is
15:37:30 gibi the current libvirt implementation agrees with the meaning I hold here and that is our biggest and most used impl
15:37:53 dansmith hence me asking you to change the actual definitions to match, for the future
15:38:26 gibi I thought you asked me to change the implementation of the libvirt driver
15:40:00 gibi btw we probably initalize the host before the libvirt version check as we ask libvirt to tell us its version to check and for that we need the host to be initalized
15:40:04 dansmith "Might be good to nail down the expectations for those two interfaces at some point." and then " I would offer to tighten up the definition of init_host() but I'd do so in a way incompatible with your definition, so probably better for me to no"
15:41:13 gibi sorry I'm still at the point where you suggested to move host.initalize to __init__. So you say we should just update the init_host interface comment to say, "do not catch up with instances"
15:41:15 dansmith ^ this is me saying that I don't think the docstrings match your interpretation of the interfaces, so if we're not going to change the code we should update the docstrings to encode the current state you want to codify
15:41:17 gibi ?
15:42:09 gibi I can change the docstring
15:44:02 gibi So I will propose a follow up with some wording in virt.driver.init_host(). Which are the other interface you see needing adjustment in the docs?
15:48:08 dansmith I just mean the split between __init__ and init_host, which have some delineation of what should be done in each, but which I clearly don't understand
15:50:07 Uggla Reminder : nova meeting in ~10mn
15:55:40 gibi dansmith: OK. Currently virt.driver.__init__ has no docstring. I will add some there too
16:04:05 Uggla #startmeeting nova
16:04:05 opendevmeet Meeting started Mon Nov 17 16:04:05 2025 UTC and is due to finish in 60 minutes. The chair is Uggla. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:04:05 opendevmeet Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:04:05 opendevmeet The meeting name has been set to 'nova'
16:04:11 Uggla Hello everyone
16:04:31 gibi o/
16:05:09 fwiesel o/
16:05:15 bauzas o.
16:05:29 bauzas I was importing the ics because I was afraid to miss the nova meeting :)
16:06:33 Uggla Let's start
16:06:47 Uggla I expect a short one
16:06:50 Uggla #topic Bugs (stuck/critical)
16:06:57 Uggla #info No Critical bug
16:07:04 Uggla #topic Gate status
16:07:15 Uggla #link https://bugs.launchpad.net/nova/+bugs?field.tag=gate-failure Nova gate bugs
16:07:18 tkajinam o/
16:07:26 Uggla #link https://etherpad.opendev.org/p/nova-ci-failures-minimal
16:07:34 Uggla #link https://zuul.openstack.org/builds?project=openstack%2Fnova&project=openstack%2Fplacement&branch=stable%2F*&branch=master&pipeline=periodic-weekly&skip=0 Nova&Placement periodic jobs status
16:07:41 Uggla #info Please look at the gate failures and file a bug report with the gate-failure tag.
16:07:48 Uggla #info Please try to provide a meaningful comment when you recheck
16:08:05 Uggla Nothing special on the gate unless I missed it
16:08:30 tobias-urdin LarsErikP: haven't heard anything about https://review.opendev.org/c/openstack/nova/+/916409 but i would like it merged as well
16:10:42 Uggla gibi, bauzas from what I read it would be good to have your point of view about ^
16:10:52 sean-k-mooney tobias-urdin: im now aware of anythign blockign that specific other then lack fo review
16:11:12 sean-k-mooney dan did have one query about fucntional tests but it was not dreictly a blocker
16:11:36 Uggla sean-k-mooney it seems Dan wants a fn reproducer
16:11:41 sean-k-mooney i dont have full context loaded on this to say if that is hard or easy to do
16:12:07 sean-k-mooney Uggla: they +1 and said they were ok to proceed if other were but yes that was tehre prefernce
16:12:47 Uggla yep, I think we can review it later, but I keep this in mind.
16:13:10 tobias-urdin sorry everyone, i was responding to a ping didn't see it was in the middle of a meeting, my bad!
16:13:53 Uggla tobias-urdin, no worries, I though it was linked more or less to gate. :)
16:15:32 Uggla skipping next point because gmaan seems not available.
16:15:40 Uggla #topic Release Planning
16:15:48 Uggla #link https://releases.openstack.org/gazpacho/schedule.html
16:15:58 Uggla #info Soft spec freeze November 20th !
16:16:25 Uggla yes this is this Thursday already.
16:16:48 bauzas as I said, I'm a bit sad about this deadline but OK
16:17:00 bauzas Uggla: can you at least send an email about this deadline ?
16:17:25 bauzas I mean, a reminder
16:17:27 Uggla yep I did it last week, but I will send an update.
16:18:10 Uggla update/reminder
16:18:35 opendevreview Merged openstack/nova master: [hacking] Improve N373 to catch also other primitives https://review.opendev.org/c/openstack/nova/+/958408
16:19:10 Uggla #topic Review priorities
16:19:23 Uggla #link https://etherpad.opendev.org/p/nova-2026.1-status

Earlier   Later